Should crosstesting work?

Ferenc Wagner wferi at
Sun Mar 2 05:51:18 CST 2003

Tony Lambregts <tony_lambregts at> writes:

>> running them in a DOS window produces the list of
>> possible tests as expected, but running a real test fails
>> with an Illegal Operation
> AFAIK This is supposed to work. I have not tried it myself
> so I can't help much. Which test(s) did you try to run? Do
> they all produce the same error?

Not exactly the same, but very similar ones.  Mostly like
this (from gdi32_crosstest generated, translated by me):

GDI32_CROSSTEST caused page fault
in module GDI32_CROSSTEST.EXE at address 014f:00455a49.
EAX=004559fc CS=014f EIP=00455a49 EFLGS=00010246
EBX=00455a20 SS=0157 ESP=0082fd7c EBP=0082fd84
ECX=c3b9ed7c DS=0157 ESI=00401220 FS=4b5f
EDX=bffbf880 ES=0157 EDI=8172b5d8 GS=0000
Bytes from CS:EIP:
64 8b 35 60 00 00 00 a1 70 80 45 00 50 e8 51 08
Stack dump:
00455a20 00401220 0082fd94 00455abd 00455a20 0095049d
0082fda4 0040145b 00401220 00000073 0082fdac 00455258
0082fdb4 00455a04 0082fdc4 00455f4a

For user32 tests I get the same (an other address with
nearly the same byte sequence is constant for every test)
with the exception of the class test, which faults at a
different address.  Btw, this is the module I put a new test
in.  That faults like the majority does. :(

I will try to build these with MSVC 6, but I seldom has
access to a computer with that software, and the process
isn't much fun altogether (according to the Developers'
Manual, which is great, thanks, Francois!)


More information about the wine-devel mailing list